This proposal establishes the statistical and data management methodology and implementation for a scientifically valid registry for the study of Alzheimer's disease (AD). It is part of a proposal by CERAD, Consortium to Establish a Registry of Alzheimer's Disease. The first component deals with the management and extension of a data base containing demographic, clinical, neurological, neuropathological and neuro-imaging data on a large group of AD patients and epidemiologic controls. The second deals with the development of standardized methods for neuro imaging and neuropathology. This includes development of standardized reporting forms and the development, description and application of valid statistical and epidemiologic methods for assessing the reliability and validity of these diagnostic and descriptive characterizations. The third deals with the development, description and application of statistical methods to (1) characterize and correlate the clinical, neuropsychologic, neuro-imaging and, where possible, neuropathologic information; (2) to investigate the potential of predicting "milestones" in the clinical history of AD, the predictors to include status at entry into the study and changes - and rates of change - over time. The "milestones" include entry into nursing home, physiological endpoints such as incontinence and death.